Monitor and manage your Python Jupyter kernels in VS Code 1.100+ or compatible Cursor. Supports Linux, Remote SSH, WSL and dev containers. No server setup needed.
Quick start
Extensions → Install from VSIX… (install on the remote host when using Remote SSH).
Click the notebook icon in the activity bar.
Expand a kernel to see its details.
Control
Use
Countdown / Refresh
Next update / update now
Status bar
Total kernel RAM; click to open the sidebar
Detail / Copy button
Copy the value
Groups: Current window → Jupyter servers → Other Kernels → Unclassified Kernels.
Kernel actions
Action
Effect
Interrupt
Request cancellation of the current computation
Stop
End the kernel; clear its in-memory state
Force Kill
End immediately, without cleanup
Use kernel buttons or the context menu. Actions require confirmation.
Settings
Search Jupyter Notebook Manager in Settings.
Setting
Description
Default
jnManager.refreshSeconds
Refresh interval (seconds)
5
jnManager.includeExternalServers
Show other local server kernels
true
jnManager.includeOtherWindows
Show other editor kernels
true
jnManager.highlights.enabled
Enable yellow warnings and red critical alerts
true
Current-window and unclassified kernels always remain visible.
Highlights
Prefix: jnManager.highlights.. Thresholds use all host CPUs and total host RAM.
Setting
Description
Default
cpuWarningPercent
🟡 CPU ≥ threshold (%)
80
cpuCriticalPercent
🔴 CPU ≥ threshold (%)
100
memoryWarningPercent
🟡 RAM > threshold (%)
80
memoryCriticalPercent
🔴 RAM > threshold (%)
90
totalMemoryWarningPercent
🟡 Status-bar total kernel RAM > threshold (%)
80
totalMemoryCriticalPercent
🔴 Status-bar total kernel RAM > threshold (%)
90
uptimeWarningHours
🟡 Process uptime > threshold (hours)
24
uptimeCriticalHours
🔴 Process uptime > threshold (hours)
48
Display
Behavior
Icons
Detail: metric severity; kernel: highest severity
Status bar
Total kernel RAM, including hidden groups; warning/error icon and background
Text
Follows explorer.decorations.colors and selection styling
Uptime warning = 0
Warns for positive process uptime
Custom colors
Set jnManager.warningForeground / jnManager.criticalForeground in workbench.colorCustomizations
Measurement limits
Metric
Measures
CPU
Kernel process usage; 100% per core; appears after the second sample
RAM
Kernel process RSS; excludes worker processes and GPU memory
Uptime
Process lifetime, not cell execution time
Only your Linux account’s kernels appear. Remote HTTP servers are unsupported. Jupyter permission is optional for monitoring.
